RELATED LITERATURE

There are different proposed methodologies to measure systemic risk and financial stability. For example, Goodhart et al. (2006) propose a general equilibrium model, which includes heterogeneous agents, endogenous defaults, and credit and deposit markets.

Additionally, there are some other works re­lated to systemic risk, which follow totally different approaches; for example, in Barnhill and Souto (2008), the authors propose to use portfolio simu­lation to study systemic risk in Brazil. In Bartram et al. (2007), the authors find that the probability of a breakdown of the international financial system is small; although, things have changed recently and their conclusions may not hold any more. In Lehar (2005), the author proposes a risk management methodology for assessing the risk in the regulators’ portfolios (financial systems); however, the author discards contagion as an important element in systemic risk. Nevertheless, as we already said, things have changed a lot in recent times.

As it was mentioned in the previous section, there are two types of contagion: direct and in­direct. Regarding direct contagion, the empirical literature on contagion through the interbank market found little evidence of contagion on their respective banking systems (Blavarg & Nimander, 2002; Boss, et al., 2004a; Boss, et al., 2006; Degryse & Nguyen, 2004; Furne, 1999; Graf, et al., 2005; Lehar, 2005; Muller, 2006; Sheldon & Maurer, 1998; Toivanen, 2009; Upper & Worms, 2004; Wells, 2002).

On a different line of research on financial contagion and systemic risk, there has been a recent furor for research on network theory (graph theory) and financial stability. Moreover, the term “too interconnected to fail” it is now part of the financial authorities’ language. It is common nowadays to see graphs and networks describing financial systems, trading networks, banking networks, interbank markets, etc. For example in Boss et al. (2004b), the authors describe the network topology of the Austrian interbank market; whereas in Iori et al. (2005), the authors study the Italian overnight market from the point of view of network theory, and Markose et al. (2009) illustrates how to develop a network model of the credit default swaps market. Despite such a current fame, there has been constant research in the past on network theory in finance and econom­ics. However, there is now the general idea that such approach could be useful in understanding, measuring and managing systemic risk (B attiston, et al., 2009). In fact, bankers are fighting to stop recent proposals to implement policies to handle the Too-Big-To-Fail (TBTF) problem; instead, they argue that such policies should be oriented to the too interconnected to fail issue⁶.

A lively discussion is taking place in global forums regarding such too interconnected to fail issue. Nevertheless, we believe that is not clear whether the network topology of the interbank market alone can be used to derive measures for fi­nancial stability or regulation purposes. Although the analytical models of Allen and Gale (2000) and some simulation models Nier et al. (2006) might disagree on this. We believe that the initial shocks, their likelihood, and the severity of the losses must be taken into account on the model­ing of systemic risk. Otherwise, the conclusions extracted might be misleading.

In our simulation model we pretend to generate macroeconomic scenarios instead of generating “shocks” to the system as the risk factors are interrelated. For example, an increase on the in­terest rates would lead to a change on the market distribution of losses and would also lead to an increase on the defaults on the credit portfolio. Along this line, banks would be affected not only on its market portfolio but on its credit portfolio as well. Additionally, the second round effects like financial contagion could cause further losses to the banks.

We contribute to the field by proposing a practical and simple approach to estimate the distribution of losses for the banking system and we show some interesting measures which can be derived once such distribution is estimated. Moreover, there is an important concern about how to measure the systemic contribution of a financial institution and with our approach it is possible to estimate the individual contribution of a particular institution to the systemic risk. Finally, our approach allows performing stress testing in a coherent way, even considering second round effects like contagion.

The remaining part of the chapter is structured as follows: Section 3 explains the simulation model in detail, Section 4 presents the results of the execution of the simulation model for Mexico, and finally, Section 5 draws some conclusions on the model and the results. Appendix A provides details on the macroeconomic model used.

3.